Hi Dylan,
Central Limit Theorem states that we can assume normality for large n--some books call large n 30, some say 50, some say 75. For this question, let's assume large n is 30 and we can apply the Central Limit Theorem.
Now, we are still working with a sample mean, which means we need standard error. The formula for this:
SE=sigma/sqrt(n) where:
sigma=population standard deviation
n=sample size
Here:
sigma=1541
n=33
SE=1541/sqrt(33)
SE=268.26
Now, we can apply the classic formula in statistics z=(x-mu)/sigma, since we are assuming normality with the central limit theorem, but we must use standard error in place of standard deviation since we are working with a sample. So:
z=(x-mu)/SE
Since we are working with two Dow gains--500 and 700--I will use x1, x2, z1, and z2 in my calculations.
z1=(x1-mu)/SE
x1=500
mu=652
SE=268.26
z1=(500-652)/268.26
z1= -0.57
From z-table, P(Z<z1)=P(X<x1)=0.2843
Keep both z1 and that probability in mind. Now let's find z2.
z2=(x2-mu)/SE
x2=700
mu=652
SE=268.26
z2=(700-652)/268.26
z2= 0.18
From z-table:
P(Z<z2)=P(X<x2)=0.5714
Now, when we have an inequality Pa<x<Pb, that means we are going to subtract the probabilities, so:
P= 0.5714 - 0.2843
P= 0.2871
High standard error lowers this probability because it ended up in a denominator. I hope this helps.